An approach to obtaining phase measurements on pulsed radio frequency (RF) data includes localizing the central line of the pulsed spectrum and processing normal vector network analyzer (VNA) measurements on that line. One disadvantage of this approach is that little of the RF power gets into the local line for low duty cycles. This is almost always the case for narrow pulses. In addition, for high profiling resolution, the duty cycle must get even smaller with the result being that dynamic range degrades. This approach is therefore largely unusable for effective duty cycles of less than 0.1%.
Another approach to obtaining phase measurements includes direct time domain acquisitions of intermediate frequency (IF) signals. This approach has been used to avoid the energy loss penalty and allow duty-cycle-independent processing. The minimum profiling resolution is set by the sample rate and the usual Discrete Fourier Transform (DFT) technique of processing collected data. The accuracy and sensitivity to noise of this approach degrades rapidly below about 8 samples so that the minimum profiling width with high accuracy is about 8 samples (i.e., 20 ns for a 400 MHz analog-to-digital converter (ADC) clock). This profiling width is too wide for many applications.